browser-use/video-use ? reverse-engineered prompt

Build me an AI video editor that I can control with chat. I want to just drop a folder of raw video takes, and then tell the agent what to do in plain English to get a final edited video back.

It should be smart enough to analyze the audio and automatically cut out all the dead air, false starts, and filler words like "um" and "uh." It should also handle basic color grading to give everything a consistent, cinematic look. I'd also like it to be able to generate and burn in subtitles directly onto the video—maybe in an uppercase, two-words-at-a-time style.

If I ask for them, it should be able to create simple text animations or overlays. The main idea is to avoid a complicated timeline editor. I just describe the final product, and it figures out the cuts and renders a clean final.mp4 for me.